Khazar Khaganate

At a glance

The borders on our map come from Cliopatria v0.12 (Seshat Global History Databank) and Historical Basemaps. They show one modelled outline per year, which is not necessarily the khaganate's historical peak.

Who were the Khazars?

The Khazars were the leading tribe of what Dan Shapira, in the Encyclopaedia Iranica entry "Khazars", calls "a polyethnic tribal confederation in Eastern Europe, from the mid-7th century until the late 10th century, led by a Turkic tribe of the same name." Shapira adds that it was possibly the most long-lived steppe power of Eurasia. Its population mixed nomads and settled peoples of Turkic, Finnic, Ugrian, Slavic, Iranian and North Caucasian origin.

The Encyclopaedia Britannica article "Khazar" stresses that the Khazar state "bore little resemblance to the other Turkic empires of central Eurasia". It had a secluded sacral ruler, towns, an economy based on taxes and tolls, and close diplomatic ties with Constantinople. Britannica also notes a key problem for historians: not a single line of the Khazar language has survived. Almost everything we know comes from Arabic, Persian, Byzantine, Armenian, Georgian, Hebrew and Old Rus' writers.

Where did the Khazars come from?

The Khazars grew out of the Western Turkic Khaganate, itself a branch of the first Türk Khaganate. Writing in the YIVO Encyclopedia of Jews in Eastern Europe, Peter B. Golden notes that references to Khazars before 630 are anachronistic, and that "it is only from about 650" that we can speak confidently of a Khazar polity. Shapira likewise argues that before the Western Turkic state fell in 659, the Khazars "cannot be distinguished from their masters, the Western Turks."

When Western Turkic rule in the west broke up, a short-lived Bulgar state briefly dominated the Pontic steppe: Old Great Bulgaria under Kubrat (Koubratos) of the Dulo clan. The Khazars destroyed it. According to Theophanes, as summarised by D. M. Dunlop in the Encyclopaedia Judaica, Kubrat's sons then went separate ways. One group stayed under Khazar rule, one went north to found Volga Bulgaria (see our article on Volga Bulgaria), and one under Asparukh crossed the Danube c. 679–681 to found the First Bulgarian Empire.

The dynasty and the name

It is unproven whether the khagans belonged to the Ashina clan, the royal house of the Türks. The tenth-century Persian geography Hudud al-'Alam makes the khagan a descendant of "Ansa", which Dunlop read as Ashina, and many scholars accept a Western Türk royal origin. The name "Khazar" itself has been derived from a Turkic verb meaning "to wander", from the Uyghur tribe Qasar, and even from the title caesar. None of these etymologies is settled.

Historians disagree: where did the Khazars come from?

Older scholarship linked the Khazars with the fifth-century Akatziri or with the Sabirs. Shapira rejects both links and treats every pre-650 "Khazar" in Arabic and Persian sources as an anachronistic label for the Western Turks. There is also an open debate over the language of the ruling tribe: was it Oghuric (Bulgar-type), as al-Istakhri implied when he wrote that Khazar resembled Bulgar, or Common Turkic? Of more than fifty known Khazar words, only Sarkel ("white fortress") is clearly Oghuric (Shapira, Encyclopaedia Iranica).

Verdict: The Khazars were a Turkic ruling tribe that emerged from the Western Türk Khaganate in the mid-seventh century and led a multi-ethnic confederation. Their exact tribal ancestry and language are unknown, and any confident claim either way goes beyond the evidence.

How did the Khazars fight the Arabs?

For almost a century the Caucasus was the front line between the Khazars and the caliphate. Arab armies of the Rashidun Caliphate raided north through Derbent (Bab al-Abwab) in 642. In 652 the Arab general 'Abd al-Rahman ibn Rabi'a was defeated and killed outside Balanjar (Shapira, citing al-Tabari; Dunlop converts the Islamic date, AH 32, to 653). This ended the first Arab–Khazar war.

The second war, against the Umayyad Caliphate, was more dangerous. The Arabs retook Derbent in 713–714, and the Khazars invaded Muslim territory in 717. In 730 they won their greatest victory: on the plain of Ardabil in Azerbaijan, after a three-day battle, the Arab army was destroyed and its commander, al-Jarrah al-Hakami, was killed. A Khazar force then raided as far as Diyarbakir and Mosul before being pushed back. Some 230 years later, King Joseph of Khazaria was still boasting of the booty taken at Ardabil.

Marwan's campaign of 737

The Umayyad general Marwan ibn Muhammad, later Caliph Marwan II, ended the war in 737. Pretending to target the Alans, he crossed the Caucasus with what the Encyclopaedia Iranica entry "Khazars" calls "an army of about 150,000", drove deep up the Volga and forced the khagan to flee. The khagan promised to accept Islam, and two jurists were sent to instruct him. Some Muslim historians record this conversion, but it was short-lived and is never heard of again. Marwan was soon drawn into the civil wars that ended Umayyad rule in 750, so the caliphate never consolidated its victory, and the Caucasus settled as the frontier.

Why were the Khazars allies of Byzantium?

The alliance with the Byzantine Empire began before the khaganate existed. In 627, according to Theophanes the Confessor, "the Turks from the East whom they call Khazars" crossed the Caspian Gates under their leader Ziebel. They joined Emperor Heraclius at the siege of Tiflis during his war against the Sasanian Empire. Ziebel is now usually identified as a ruler of the Western Türks rather than a Khazar khagan, so the episode belongs to Khazar prehistory rather than Khazar history proper.

Under the khaganate itself, the alliance was sealed by marriages. The exiled Justinian II married the khagan's sister around 704; she took the baptismal name Theodora. In 732 Leo III married his son, the future Constantine V, to the khagan's daughter Tzitzak ("flower" in Turkic), who was baptised Irene. Their son reigned as Leo IV "the Khazar" (775–780). From the later seventh century the Khazars held most of the Crimea outside Byzantine Cherson, and they also helped Byzantium there.

Sarkel: a Byzantine-built fortress

The clearest proof of cooperation is Sarkel on the lower Don. In chapter 42 of De Administrando Imperio, Constantine VII Porphyrogenitus relates that the khagan and the bek asked Emperor Theophilos (829–842) to build them a fortress. The official Petronas Kamateros supervised the work with Byzantine builders. The date is debated – c. 833 in older literature, while Peter B. Golden dates the building to 840–841. Sarkel's name means "white fortress"; Old Rus' sources call it Bela Vezha.

How was Khazaria governed?

Khazaria had a dual kingship. The khagan was a sacral figure who embodied the state's heavenly fortune, but by the tenth century he held little practical power. The bek (beg, ishad) led the army and ran the government. Ibn Rusta, writing c. 903, says the supreme ruler does not exercise any real power. Al-Istakhri (c. 932) adds the famous, perhaps exaggerated story that a new khagan was half-throttled with a silk cord and asked how many years he wished to reign.

The khagan's sacral seclusion seems to have developed gradually. In 862 Methodius, brother of Constantine-Cyril, still described the khagan as an active ruler with a chief adviser close to him, presumably the bek. Historians have often linked the rise of the bek to the conversion to Judaism, but Shapira notes that such views "cannot be substantiated". Golden has suggested that the growing power of a Muslim guard of Khwarazmian origin mattered more.

Seven judges and a Muslim guard

In The Meadows of Gold, written in the 940s, al-Mas'udi describes the legal pluralism of Itil. In the translation by Paul Lunde and Caroline Stone: "It is a rigid custom in the Khazar capital that there should be seven judges: two for the Muslims; two for the Khazars, who make their decisions in accordance with the Torah; two for the Christians, who make theirs according to the Gospels; and one for the Slavs, Rus and other pagans." Mas'udi adds that the king "is the only ruler of these eastern countries to have a paid army". That army was made up of Muslims called the Arsiyya, who had migrated from Khwarazm.

Itil itself was a twin city. Arabic geographers describe a western part with the royal palace, which stood on an island, and an eastern commercial quarter (Khazaran). Muslims, Jews, Christians and pagans lived there side by side, each with their own places of worship.

Why did the Khazars convert to Judaism?

The traditional story comes from the Khazar Correspondence, an exchange of Hebrew letters from c. the 950s–960s. They passed between Hasdai ibn Shaprut, the Jewish courtier of the Caliph of Córdoba, and Joseph, king of the Khazars. Joseph's Reply tells how his ancestor Bulan was visited by an angel in dreams, plundered the caliphate's lands and built a sanctuary. Bulan then held a debate between a Jew, a Christian and a Muslim, after which he and his nobles chose "the religion of Israel". A later king, Obadiah, allegedly brought in sages and introduced the Mishnah, the Talmud and synagogue liturgy.

A second Hebrew text tells a different story. This is the Cambridge or Schechter Document from the Cairo Genizah (c. 949), published in English by Norman Golb and Omeljan Pritsak in Khazarian Hebrew Documents of the Tenth Century (1982). In its version, Jews already lived among the Khazars without the Torah; a Jewish general returned to his faith, a religious debate followed, and books were found in a cave. Its hero receives the Hebrew name Sabriel.

Why Judaism?

Historians usually explain the choice in political terms. Becoming Christian would have implied deference to the emperor in Constantinople, and becoming Muslim deference to the caliph in Baghdad. Judaism gave the Khazar rulers a monotheism with scripture but without a foreign overlord. Shapira adds a cultural angle. Khazar Judaism grew up in a milieu shaped by Islamic culture and expressed the empire's claim to be a "Third Force". Meanwhile, Jewish merchants and refugees from Byzantium and the Islamic lands brought the faith to the Khazar towns.

What the outside sources say

Contemporary Muslim writers confirm that the ruling house was Jewish. Ibn al-Faqih (c. 903) reported that the Khazars were Jews and had become so only recently. Ibn Rusta (c. 903) said the supreme ruler and the bek were Jewish, but that Khazar paganism resembled that of the Turks. Al-Istakhri (c. 932) said the king and his court were Jews, but that Jews were the smallest religious group. Al-Mas'udi dated the king's conversion to the caliphate of Harun al-Rashid (786–809).

Under 986, the Russian Primary Chronicle describes "the Jewish Khazars" offering their faith to Vladimir the Great. The episode is a literary one, written long afterwards, but it shows that Kievan writers associated the Khazars with Judaism.

The "Moses" coins

The strongest material evidence comes from coins. In 2002 a dirham from the Spillings hoard on Gotland was identified as Khazar; the hoard, with 14,295 coins, is the largest Viking-age silver hoard. The coin was minted in 837/838 as an imitation of Abbasid coinage, but it replaces the Muslim formula with the Arabic phrase "There is no god but God and Moses is the messenger of God" (Roman Kovalev's studies, summarised by Shapira). A handful of related coins are known, and Golden also points to Khazar imitation dirhams inscribed ard al-Khazar, "land of the Khazars".

Historians disagree: when, and how far, did Judaism spread?

Judah Halevi's Kuzari (c. 1140) puts the conversion 400 years earlier, i.e. c. 740. Dunlop (1954) and others proposed two stages: a simple biblical faith from c. 740, then rabbinic Judaism c. 800. Constantin Zuckerman ("On the Date of the Khazars' Conversion to Judaism", Revue des études byzantines, 1995) linked the conversion directly to the failure of Constantine-Cyril's Byzantine mission c. 861. Shapira places the royal conversion in the ninth century, pointing out that the Georgian Martyrdom of St Abo (786–787) still describes the Khazars as pagans. Shaul Stampfer ("Did the Khazars Convert to Judaism?", Jewish Social Studies 19/3, 2013) went furthest. He argued that some of the sources are "pseudepigraphic" and the rest "of questionable reliability", and concluded "that such a conversion never took place". Shapira calls this view "unconvincing".

Verdict: A Jewish ruling house in ninth- and tenth-century Khazaria is well supported by independent Muslim writers and by the "Moses" coins of 837/838. The date of c. 740 rests mainly on a twelfth-century literary work and is weaker; a ninth-century date (c. 800–861) fits the evidence best. Judaism was an elite religion – Muslim observers called Jews the smallest group – and there is no evidence that the population as a whole converted.

A tolerant, plural society

Khazaria never imposed Judaism on its subjects. Christians had bishops in the Crimea and the Caucasus, Muslims had mosques, their own judges and places in the royal guard, and many steppe tribes kept the cult of Tengri. Around 861 Constantine-Cyril, the future apostle of the Slavs, led a Byzantine embassy to the khagan and debated with Jewish and Muslim scholars. According to his Life, the mission baptised some people but did not convert the court.

Judah Halevi later used the conversion story as the frame for his philosophical dialogue, the Kuzari, whose Arabic title means roughly "The Book of Refutation and Proof in Defence of the Despised Faith". In it, a Khazar king questions a philosopher, a Christian, a Muslim and a rabbi. It is a work of literature, not a historical source for Khazaria.

How rich was Khazaria? Trade, tribute and tolls

Khazaria's wealth rested on livestock, gardening and fishing, and above all on control of trade routes. Shapira lists the goods that passed through: furs, salted and smoked fish, candle wax, honey, jewellery, silverware, isinglass and Islamic coins. Slaves – pagans captured in the north – were sold to the Muslim world. The state taxed merchants and ships at Itil, which sat where the Volga trade met the routes to Khwarazm, the Caspian and Byzantium.

Kovalev notes that millions of Islamic silver dirhams flowed north along the "Khazar Way" (lower Volga–Don–Donets–Oka–upper Volga) to north-western Russia and Scandinavia, in exchange for Rus' furs and slaves. This is the trade that brought Scandinavians – the Varangians – to the rivers of Eastern Europe. The ninth-century Persian geographer Ibn Khordadbeh also describes Jewish "Radhanite" merchants travelling between Europe and China. Whether they used the Khazar route is uncertain, and Shapira stresses that evidence for Jewish traders in Khazaria itself is scarce.

Tribute from the Slavs and the Bulgars

According to the Russian Primary Chronicle (translated by Samuel H. Cross and Olgerd P. Sherbowitz-Wetzor, 1953), the Khazars imposed tribute on several East Slavic tribes. The Polianians around Kiev once paid "one sword per hearth". Under the year 859, the Polianians, Severians and Viatichians gave "a white squirrel-skin from each hearth". The Radimichians paid "a shilling apiece" until 885, when Oleg of Novgorod took their tribute for himself. The Chronicle's early dates are artificial, but the pattern of Khazar overlordship along the Dnieper and Oka is widely accepted.

Volga Bulgaria was also a Khazar tributary. In 921–922 the Abbasid envoy Ahmad ibn Fadlan travelled to the Volga Bulgars, whose ruler was seeking the caliph's support to escape dependence on the Khazars. Ibn Fadlan's account also contains one of the fullest descriptions of the khagan's sacral seclusion; see our article on Ibn Fadlan's Risala.

What has archaeology found?

Archaeologists associate the Khazar period with the Saltovo-Mayaki culture of the eighth to tenth centuries, found across the Don and Donets basins, the Azov region and the Crimea. It combines hill-forts, unfortified villages, large cemeteries, stone-built fortresses and wheel-made pottery, and reflects the mixed Alan, Bulgar and Turkic population of the khaganate rather than a single "Khazar" people. Svetlana Pletnyova, who took part in the Sarkel excavations, became its leading interpreter.

Sarkel

Mikhail Artamonov excavated Sarkel in the 1930s and again in 1949–1951. He uncovered a rectangular brick fortress, Byzantine architectural material and a later Rus' settlement. The site was then flooded in the early 1950s, when the Tsimlyansk reservoir was filled, and can no longer be excavated on land. Artamonov's Istoriya khazar (1962) remains a classic. Shapira notes, however, that Soviet scholarship – Artamonov included – increasingly portrayed the Khazars negatively after a 1951 Pravda article that he attributes to Stalin.

The search for Itil

Itil has never been securely located. Some scholars believed it was drowned by a rise in the level of the Caspian. In 2008 Dmitry Vasilyev of Astrakhan State University announced that the lowest layers of the Samosdelka site, south-west of Astrakhan, contained a triangular brick fortress and yurt-like dwellings, and identified the site as Itil.

Historians disagree: has Itil been found?

Vasilyev's team argued that the brick fortress, imported goods and traces of a great fire at Samosdelka fit the written descriptions of Itil, where only the king was allowed to build in brick. Critics disagree. The editor of a 2018 issue of Russian Studies in History devoted to the Khazars called the claim possibly "premature", since Samosdelka lacks eighth- and ninth-century Khazar layers like those at Sarkel. A Russian Geographical Society report of 15 June 2020, "On the Threshold of the Great Discovery", noted that Saltovo-Mayaki artefacts "were not found" at Samosdelka and that work had shifted to the Semibugry site.

Verdict: Samosdelka is a genuine Khazar-era and post-Khazar settlement in the Volga delta, but its identification with Itil is unproven. Treat Itil's exact location as unknown.

How big was Khazaria?

Khazaria's borders were never fixed lines. Its core lay between the lower Volga, the lower Don and the northern Caucasus, with the Crimea, the Kuban and Daghestan as southern and western provinces. Kovalev describes Khazar authority at its widest, in the first half of the ninth century, as reaching from the middle Dnieper in the west to the Volga–Ural steppe in the east, and from the middle Volga in the north to the Crimea in the south. That is roughly today's southern Russia, eastern Ukraine, western Kazakhstan and northern Azerbaijan.

Much of this was a zone of tribute rather than direct rule. The Magyars lived under Khazar suzerainty in the ninth century, Volga Bulgaria and East Slavic tribes paid tribute, and the Caucasian principalities shifted between Khazar and caliphal influence. No reliable figures for the khaganate's area exist, and the outline on our map should be read as a model rather than a measured frontier.

Why did the Khazar Khaganate fall?

Khazaria declined in stages. From the 890s, Pechenegs and Oghuz pushed into the steppe from the east, and the Magyars left the Khazar orbit for the Carpathian Basin. The Rus' annexed Khazar tributaries along the Dnieper, while the Volga trade increasingly bypassed Khazar toll collectors in favour of Volga Bulgaria. Byzantium, once an ally, turned hostile under Romanos I Lekapenos (920–944) after the Khazars retaliated for Byzantine persecution of Jews, and it encouraged the Rus' and the Alans against Khazaria.

The Cambridge Document describes a war in the 940s between the Khazars and a Rus' leader called HLGW (probably Helgi, i.e. Oleg), who was then forced to attack Byzantium. These events connect with the failed Rus' attack on Constantinople in 941 under Igor of Kiev. Under Olga of Kiev, Kievan power consolidated, and her son Sviatoslav finished what his father had begun.

Sviatoslav and the destruction of Itil

Under the year 965 the Primary Chronicle reports that "Svyatoslav sallied forth against the Khazars", defeated their army led by "their Prince, the Kagan", and "took their city of Bela Vezha" – that is, Sarkel (see our biography of Sviatoslav I of Kiev). The Arab geographer Ibn Hawqal, who was in the region, dates the Rus' sack of Bulgar, Khazaran, Samandar and Itil to AH 358, i.e. 968/969. He wrote that "nothing remains" of the Bulgars, Burtas and Khazars. Ibn Miskawayh separately records an attack by the Oghuz in AH 354 (965).

Most historians now reconstruct a campaign, or series of campaigns, between 965 and 969. In it, Rus' forces with Oghuz and perhaps Pecheneg allies took Sarkel and then destroyed the cities of the lower Volga. Shapira argues that 968/969 is the decisive date and that Sviatoslav's victory was "Pyrrhic": he soon turned to the Danube, leaving the steppe to the Oghuz and Pechenegs.

Remnants after 969

Khazar refugees returned to Itil under the protection of the Shirvanshah or of Khwarazm, reportedly on condition that they accept Islam. The Khazar name survived in the Crimea. In 1016, according to Kedrenos and Skylitzes, a joint Byzantine–Rus' fleet captured "Georgius Tzul", called khagan, at Kerch. The Encyclopedia of Ukraine dates the last chronicle mention of Khazars to 1079, in Tmutarakan, and an eleventh- or twelfth-century seal calls Oleg of Tmutarakan "archon of all Khazaria". By the twelfth century the steppe belonged to the Cuman-Kipchaks.

What was the legacy of the Khazars?

The Khazars shaped early Rus'. The Rus' rulers of Kiev adopted the title kagan, and Metropolitan Hilarion's mid-eleventh-century Sermon on Law and Grace praises Vladimir as "the great Kagan of our land" (English text as published by Durham University). The tribute system that Kievan Rus' inherited had been organised under Khazar overlordship; see also our Kievan Rus' glossary entry.

In Hungary, the Kabars – three Khazar tribes who rebelled and joined the Magyars, according to chapter 39 of De Administrando Imperio – became part of the people who founded the Principality of Hungary. As late as 1309 a Hungarian church council forbade Catholics to marry people described as Khazars. Claims that the Crimean Karaites or Krymchaks descend from the Khazars are unproven. Shapira states that all the "Khazar evidence" claimed for Crimean Jewish graves is forged, and he has shown that documents produced by Abraham Firkovich in the nineteenth century are forgeries.

The Caspian Sea is still called the "Khazar Sea" in Persian, Turkish and Azerbaijani. The conversion story lives on in literature, from the Kuzari to Milorad Pavić's novel Dictionary of the Khazars (1984).

Are Ashkenazi Jews descended from the Khazars?

No. Genetic and historical evidence shows that Ashkenazi Jews descend mainly from a mixture of Middle Eastern (Levantine) and southern and western European ancestry, with no detectable Khazar contribution. The idea of a Khazar origin was popularised by Arthur Koestler's book The Thirteenth Tribe (1976). Shapira traces it to a national myth among Hungarian Jews in the late Habsburg period, which paired a Khazar origin for them with the steppe origin claimed for the Magyars.

What the genetic studies show

The largest test of the hypothesis is by Doron M. Behar and colleagues: "No evidence from genome-wide data of a Khazar origin for the Ashkenazi Jews" (Human Biology 85/6, 2013). It analysed 1,774 samples from 106 populations, including 261 samples from 15 populations of the Caucasus and the region directly to its north. The authors found "No particular similarity of Ashkenazi Jews with populations from the Caucasus", "particularly with the populations that most closely represent the Khazar region". Ashkenazi Jews share most of their ancestry with other Jewish and Middle Eastern populations, plus European admixture.

Ancient DNA has since confirmed the picture. Shamam Waldman and colleagues (Cell 185, 2022) analysed the genomes of 33 Jews buried in fourteenth-century Erfurt, Germany. They were genetically similar to modern Ashkenazim and showed an extreme "founder event" that predated the fourteenth century, with little gene flow into the population afterwards. Their ancestry combined Middle Eastern and European sources. That is the pattern expected from Jewish migration through the Mediterranean and the Rhineland, not from the Volga steppe.

The dissenting view and its critics

The main scientific dissent is by Eran Elhaik: "The missing link of Jewish European ancestry: contrasting the Rhineland and the Khazarian hypotheses" (Genome Biology and Evolution 5/1, 2013). Using 1,287 individuals from 8 Jewish and 74 non-Jewish populations, Elhaik read the genetic similarity between Ashkenazim and Caucasus groups such as Armenians and Georgians as evidence of "Judeo-Khazarian" ancestry. Critics noted three problems. Modern Armenians and Georgians are poor stand-ins for the Turkic-led Khazars; Caucasus and Near Eastern ancestry overlap heavily; and, as the Jewish Review of Books pointed out, the study contained no Khazar DNA at all.

A 2019 bioRxiv preprint by Mikheyev and colleagues analysed "whole-genome sequences of eight men and one woman buried within the distinctive kurgans of the Khazar upper (warrior) class" in the Rostov region. It found a polyethnic elite of mainly Central Asian ancestry that "left few persistent genetic traces in Europe". The study is not peer-reviewed and the identification of its samples has been criticised, so it should be treated as preliminary.

How the myth is misused

The Khazar question became politically charged because antisemites use it to claim that Jews are "not real Jews" and have no historical connection to the Land of Israel. The American Jewish Committee's Translate Hate glossary explains how the trope is used to deny that connection. It cites a 2023 speech in which Mahmoud Abbas claimed that European Jews descended from "the Khazar Kingdom". The Canadian Anti-Hate Network documents how the myth spreads in conspiracy communities.

These uses are false on every level. Some Khazars may have joined later Jewish communities; the YIVO Encyclopedia says other Jewish communities in the region "undoubtedly absorbed" Khazar Jews and Judaised Khazars. Even so, this would not change the documented Middle Eastern ancestry of Ashkenazi Jews, and Jewish identity has never depended on genetic purity. The history of a medieval steppe empire offers no support for modern conspiracy theories.

Timeline

What the map shows next

On our map the Khazar Khaganate is drawn from AD 630 on the map to AD 979 on the map. In 805 it still dominates the steppe between the Avars in the west and the Oghuz and Kimek-Kipchaks in the east. Step forward to AD 965 on the map for the eve of Sviatoslav's campaign and to AD 980 on the map for the power vacuum left by the khaganate's fall. By 1000 on the map the steppe is divided between Kievan Rus', the Pechenegs and Volga Bulgaria, with the Samanids and Abbasids to the south-east.